- Email: [email protected]
Disturbances in time estimation during absence seizures in children
Epilepsy Res., 9 (1991) 148-153 Elsevier
148
EPIRES 00418
Disturbances in time estimation during absence seizures in children
E.L.J.M. van Luijtelaar a, S.F.T.M. de Bruijn a, A.C. Declerck b, W.O. Renier c, J.M.H. Vossen a and A.M.L. Coenen a a Department of Psychology, University of Nijmegen, Nijmegen (The Netherlands), ~ Epilepsy Center Kempenhaeghe, Heeze (The Netherlands) and c Department of Child Neurology, University Hospital Nijmegen, Nijmegen (The Netherlands) (Received 28 January 1991; revision received 29 April 1991; accepted 13 May 1991)
Key words: Spike wave discharges; Absence epilepsy; Cognition; Performance; Timing behavior
In the present study children suffering from primary generalized absence epilepsy were asked to estimate time under EEG monitoring. They were asked to press a button when they thought that a fixed period of time had elapsed. Only the first response after the passage of the interval was reinforced. The dependent variable was the duration of the interval between the start of the trial and the first response, the post-reinforcement time. This parameter was used as an index for the accuracy of time estimation. The performance of the subjects in trials with and without spike wave discharges was compared. Short spike wave discharges ( < 3 s) prolonged the duration of the postreinforcement pause while longer ones reduced its duration. The prolongation was longer than could be anticipated from the duration of the spike wave discharge. It was also found that the time between the end of an EEG paroxysm and the first response was significantly shorter in trials with long spike wave discharges than in trials with short spike wave discharges. There were no differences between subjects with and without spike wave discharges. It was concluded that this type of time estimation task is sensitive in detecting cognitive disturbances induced by both short and long spike wave discharges. Moreover, it seems that after long spike wave discharges patients behave differently and are perhaps more severely disturbed. The duration of the misperception of the interval after the short spike wave discharges is larger than could be predicted from the duration of the paroxysms, which seems in contrast with the view that cognitive disturbances are only limited to the period of epileptic EEG phenomena.
INTRODUCTION
w i t h o u t evident changes in clinical b e h a v i o r 1. Brief lapses in the p a t i e n t ' s ability to m a i n t a i n c o n -
In m a n , the a b n o r m a l b r a i n activity characteristic o f
tact with the e n v i r o n m e n t are the p r i m a r y s y m p t o m o f
v a r i o u s types o f epilepsy is generally a c c o m p a n i e d by
generalized absence epilepsy. Various indications h a v e
an i m p a i r m e n t o f cognitive functioning. This i m p a i r m e n t a l processes to a c o m p l e t e arrest o f c o g n i t i v e ac-
been assigned to this s y m p t o m such as a loss o f c o n sciousness or a r e d u c t i o n in the ability to process i n f o r m a t i o n . Increases in reaction times 9A°,22, in omis-
tivity. T h e kind o f cognitive deficit is highly d e p e n d e n t
sions in c o n t i n u o u s p e r f o r m a n c e tasks as well as in
on the type, the d u r a t i o n a n d the n a t u r e o f the epileptic
stimulus detection are f r e q u e n t l y r e p o r t e d to a c c o m p a n y the characteristic spike w a v e discharges 15.
m e n t ranges f r o m a h a r d l y noticeable interference with
discharges and o f the r e q u i r e m e n t s o f the task 4. Subtle changes in m e n t a l f u n c t i o n i n g m a y even be present
It is agreed that s o m e tests are m o r e sensitive for c o g n i t i v e disturbances
Correspondence to: Dr. E.L.J.M. van Luijtelaar, Department of Psychology, University of Nijmegen, P.O. Box 9104, 6500 HE Nijmegen, The Netherlands.
t h a n others.
Simple m o t o r
tasks, such as r h y t h m i c t a p p i n g , simple r e a c t i o n t i m e or t r a c k i n g tasks are relatively little a f f e c t e d by generalized spike w a v e activity whereas m o r e c o m p l e x
0920-1211/91/$03.50 © 1991 Elsevier Science Publishers B.V. All rights reserved
149 tasks such as choice reaction time tasks, signal detection and short-term memory tasks are more sensitive 6,8,9,11,12,18,21,23,24. The sensitivity of tasks for cognitive impairments is especially important in the case of EEG discharges, since they often remain unnoticed by the patient or his environment. Furthermore, it is sometimes felt that cognitive disturbances are not present during short EEG paroxysms 12,19. In the present experiment, data will be presented which show that performance deficits do occur also during short-lasting spike wave activity. The task is a simple free operant learning task in which subjects had to estimate the duration of a fixed period of time. The response pattern induced by this fixed interval (FI) task is characterized by a long period without responding (the so-called post-reinforcement pause, initiated by the start of a new trial and ended by the first response of the subject in that trial). This task has been found to be sensitive for the presence of spike wave discharges in an animal model for absence epilepsy 27. METHODS
Subjects Fourteen subjects were recruited from 2 outpatient epilepsy clinics in the south-east of The Netherlands. All patients had a confirmed diagnosis of primary generalized absence epilepsy and their EEG had recently ( < 2 years) shown spontaneously occurring bilateral synchronous generalized 3-Hz spike wave activity with essentially normal interictal electrical activity. During the task hour, only 5 patients showed bilaterally appearing 3-Hz spike wave discharges. The ages of these 5 children ranged between 6 and 17 years, with a mean of 12. All patients were on mono- or polytherapy anti-epileptic drugs. Material The test device consisted of a multicolored wooden box (41 x 41 x 25 cm) containing a rotating disk (31.5 cm diameter), a non-transparent lid on top, and a push button. An opening in the front panel allowed collecting a token (pick-up tray). The box was made at the Instrumentation Department of the Psychology Department. The steerable slowly rotating disk in the box contained 8 compartments. A Playmobil toy could be placed in one of these compartments. Each compartment could fold down, and the toy would then fall
into the pick-up tray. An Apple II computer controlled the start and stop of the disk and controlled the presentation of the toy. During the experiment EEG traces (F0-C0, F4-C4, F3-C3, P3-O1, International 1 0 - 2 0 system) were registered, in which signals of 1 - 70 Hz were allowed to pass. The EEG was recorded on a cassette of a Medilog system, together with the responses of the subjects, the presentation of the reinforcers and the beginning and the end of the intervals. Later, all registrations were written out on paper.
Procedure After the EEG set-up was made and checked, subjects sat in front of the test box. They were shown the silently rotating disk, with its speed of 1 rpm, although this speed was not mentioned to the subjects. The toy was placed on the front compartment of the rotating disk. Subjects were shown that pressing the button had no consequences before the disk had made a complete turn, i.e., during the first minute. However, the first button press after completion of the turn stopped the rotating and the part of the disk on which the toy was placed folded downwards. The toy fell into the pick-up tray and could be collected by the subject. Next, the experimenter replaced the toy and a second trial was initiated by the experimenter who announced the start of the next trial by presenting an auditory stimulus. The dependent variable was the post-reinforcement pause, i.e., the interval between the auditory stimulus and the subject's first response. Subjects were allowed to get acquainted with the system, they were trained during 5 - l0 min. Next, the box was closed and the sign for the next trial was presented. Subjects were instructed to estimate the time of a complete turn of the disk. If they felt that this had happened, they pressed the button. In each trial the first response after the passage of 1 min was reinforced by the toy. Earlier bar presses had no consequences. The total session lasted 1 h. The collected toys were exchanged for Dfl 5, - in total at the end of the session for all subjects. The EEG was analyzed by an expert neurophysiologist who was unaware of the task requirements and behavioral data (A.D.). Each subject served as his/her own control, which means that possible differences between subjects with respect to medication are not critical, t-Tests for correlated means were used in order to test differences between
150 trials with and without spike wave discharges, the distribution free sign test was used to test differences in response latencies and a t-test for independent groups was used to test differences between subjects with and without spike wave discharges.
TABLE I The post-reinforcement pause (s) o f trials without, with short (< 3 s) and with long (> 3 s)spike wave discharges o f 5 children showing spike wave discharges in the EEG n
=
number of trials
Subject Without
n
Short
n
Long
V.M. S.H. J.S. C.B. M.B.
58.7 46.4 56.9 25.6 69.9
34 35 42 29 30
82.9 52.3 59.8 28.8 91.7
3 9 5 8 3
47.1 41.1 43.2 31.8 61.1
Mean S.D.
51.5 16.7
n
RESULTS Nine patients had no spike wave discharges during the experiment. The mean and standard deviation o f the post-reinforcement pause was 51.9 _+ 9.8 s, the number of responses per trial was 2.5 ± 0.9 and the number of trials per hour was 51.8 + 2.2. Five of the 14 subjects showed the generalized 3-Hz spike wave discharges in the E E G during the task. Their mean number of E E G paroxysms was 8.4 ± 3.1. The mean o f the number o f responses per trial was 2.6 ± 1.7 and the number o f trials per hour was 44.4 ± 5.3. There was no statistical difference in the number o f responses per trial between subjects with and without spike wave discharges. All other comparisons were within the subjects with spike wave discharges. A distinction between short ( < 3 s) and long ( > 3 s) spike wave discharges was made. Table I shows differential and opposite effects for trials with and without spike wave discharges. The post-reinforcement pause tended to be longer in trials with short spike wave discharges than in trials without spike wave activity (t -- 2.24, d f = 4, P < 0.10). A d ditionally, the post-reinforcement pause tended to be shorter after long spike wave discharges than after trials without spike wave activity (t = 2.46, d f = 4, P < 0.10). The difference between the post-reinforcement pause in trials with short and long spike wave discharges was significant (t -- 2.87, d f = 4, P < 0.05). F o r each subject the duration o f the E E G paroxysms was determined and c o m p a r e d with the prolongation o f the post-reinforcement pause. The prolongation o f the post-reinforcement pause (11.5 ± 10.6 s) was significantly longer (t = 4.24, d f = 4, P < 0.05) than the duration o f the spike wave activity (1.5 ± 0.6 s). Response latencies (time between the end o f the spike wave discharge and the first response) are presented in Table II. The non-parametric sign test showed that the latencies were significantly longer after short- than after long-lasting paroxysms (P < 0.05).
63.1 25.1
44.9* 10.7
* P < 0.05 (short vs. long).
TABLE II Response latencies (s) after short (< 3 s) and long (> 3 s) spike wave activity f o r the 5 children with spike wave discharges in the EEG during the test interval
t = time between the end of epileptic activity and the first response; n = number of spike wave discharges. Subject
t
t
short
n
long
n
V.M. S.H. J.S. C.B. M.B.
44.0 18.0 31.9 17.9 53.9
3 9 5 8 3
9.6 4.5 1.0 8.1 31.7
5 2 1 4 2
Mean S.D.
33.1 15.9
11.0" 12.1
* P < 0.05.
DISCUSSION All subjects p e r f o r m e d reasonably well in the fixed interval task. All, except the youngest patient, showed low-rate responding with only 1 - 2 responses in total at the end o f the interval. This behavioral pattern is c o m m o n l y found in h u m a n learning 2,14,2°. This means that, despite the lack o f extensive fixed interval training, the instruction and a short training were successful in inducing characteristic fixed interval responding. O f interest is also the similar number o f responses per trial o f the 5 subjects in trials without spike wave activity and o f the 9 subjects without any
151 sign of aberrant EEG activity. This suggests that interictal performance is not affected. Only 5 of the 14 subjects showed spike wave discharges during the experimental hour. It is likely that the mental activity during the task influenced and decreased the number of spike wave discharges 1. The duration of the post-reinforcement pauses shows that spike wave activity does affect the estimation of a time interval in all patients. Short spike wave discharges prolong the duration of the post-reinforcement pause and long spike-wave discharges reduce its duration. It is admitted that the distinction between short and long spike wave discharges is not solidly based on theoretical grounds. It would be preferable to make a distribution of the lengths of the spike wave discharges: a bimodal distribution with clear cut-off points would justify the distinction. However, in other studies the same distinction between short and long spike wave discharges has also been made 12. In some of them, cognitive aberrations were only present if the EEG paroxysm exceeded 3 s7'12. Only few papers can be found in which procedures or tasks are described which are sensitive for these short attacks 7,24. In this study, short spike wave discharges induce an increase of the post-reinforcement pause. Furthermore, this increase is larger than the duration of the spike wave discharge. It is often felt that the cognitive disturbances are limited to the period of the epileptic EEG phenomena and do not extend beyond the absence periods. Here, the amount of time that is wiped out or missed by the subjects is longer than the actual duration of the EEG paroxysm. It is not clear whether this 'wipe-out period' should be considered to be genuine amnesia, and if so, whether it is retro- or anterograde amnesia. Retrograde amnesia after discharges larger than 3 s has been described 6,12, whereas Hutt and Gilbert I 1 found retrograde amnesia in a short-term memory task. So far, anterograde memory deficits have not been reported. The present data do not allow us to draw conclusions with respect to putative memory deficits which extend the duration of the EEG signs. It is of interest that Mirsky and Grady 16 found a significant increase in the power of several EEG frequency bands as early as 20 s preceding the epileptic bursts, monotonically increasing till cumulating in the fully developed spike wave burst. This allows the possibility that the long cognitive deficits found in the present work are related to elec-
trophysiological antecedents preceding the fully developed spike wave discharge. In addition to this, Brunia 5 found indications for an altered motor system immediately after spike wave discharges. This latter result may indicate that antecedents of spike wave complexes are present after the EEG phenomena have disappeared. Long spike wave discharges induce an opposite effect on the post-reinforcement pause: it is shorter and relatively fast responding occurs after the paroxysmal activity has stopped. The response latency was significantly shorter than after short spike wave discharges. Thus it seems that after these long spike wave discharges a disorientation in time has taken place: the subjects seem to have no idea how much time has elapsed since the beginning of the trial and he/she responds quickly. It is well-known that after a long attack a patient's train of thought is often interrupted in such a way that he has to regain his thoughts after the attack. However, the patient will regularly carry on with his pre-ictal activities as if nothing has happened or may resume speaking or counting where he left off 17. This will occur with relatively simple tasks. Obviously, the time estimation task is sensitive to epileptic activity in such a way that a long attack interrupts the cognitive functioning so severely that disorientation and fast responding occurs after the attack. Although all subjects received the same instruction and sat in front of the same apparatus, it is not clear how subjects mastered this prospective time estimation task. Several strategies can be followed to accomplish this task: some subjects may count while others are really estimating the time that has elapsed and others may visualize the rotating of the disk. Although the type and nature of the cognitive strategy are of interest, the outcome of the present study is that similar effects with all 5 subjects were obtained, suggesting that the type of cognitive activity or the strategy used by the subjects did not have a major effect on performance disturbances induced by the spike wave activity. It is of interest that in rats with spontaneously occurring spike wave discharges the post-reinforcement pause in a fixed interval task was also prolonged in trials with spike wave discharges compared to trials without spike wave discharges 27. Also in agreement with the present study it was found that the prolongation of the post-reinforcement pause was larger than the duration of the EEG paroxysms. In the rat study,
152 a distinction between short and long spike wave
pected f r o m the duration o f the epileptic activity. Long
discharges was not made, but the agreement between
spike wave discharges disrupt ongoing activity even
the short-lasting spike wave discharges in m a n and rat
more severely, the subjects seem time disoriented.
is striking. This agreement strongly supports the validity o f this rat strain as a model for absence epilepsy in m an 25,26. It can be concluded that the timing task used in the
ACKNOWLEDGEMENTS We are indebted to Dr. J. Arends (Janssen Phar-
present experiment is sensitive in detecting cognitive
maceutics),
dysfunctioning.
In all subjects short spike wave
(Kempenhaeghe) for their cooperation in this study.
discharges prolonged the post-reinforcement pause
We also received assistance from the staff members of
and this prolongation exceeded the duration of the
the E E G Department of Kempenhaeghe, Mrs. Inge
E E G paroxysms. This relatively long 'missed' period
van der Linden and Mrs. Loan Kho, and secretarial
Dr.
L.
Oei
and
Dr.
J.
Doelman
is in contrast with the general opinion that cognitive
help f r o m Mrs. Tilly de Leijer. The study was made
disturbances are limited to the period of, in any case short, epileptic activity. We found that the magnitude
possible by support f r o m the Dutch Epilepsy Foundation 'D e Macht van bet Kleine'. Dr. E . L . J . M . van
o f the change in cognitive behavior after short spike
Luijtelaar is supported by the Royal Dutch A cad em y
wave paroxysms was much larger than could be ex-
of Sciences.
REFERENCES
12 Jus, A. and Jus, K., Retrograde amnesia in petit mal, Arch. Gen. Psychiatry, 6 (1962) 163 - 167. 13 Kasteleijn-Nolst Trenit6, D.G.A., Bakker, D.J., Binnie, C.D., Buerman, A. and van Raaij, M., Psychological effects of subclinical epileptiform EEG discharges: scholastic skills, Epilepsy Res., 2 (1988) l l l - 116. 14 Lowe, C.F., Determinants of human operant behavior. In: M.D. Zeiler and P. Harzem (Eds.), Advances in Analysis of Behavior." Vol. 1. Reinforcement and the Organisation of Behaviour, Wiley, Chichester, 1979, pp. 159- 192. 15 Mirsky, A.F., Information processing in petit mal epilepsy. In: B.P. Hermann and M. Seidenberg (Eds.), Childhood Epilepsies." Neurophysiological, Psychosocial and Intervention Aspects, Wiley, Chichester, 1989, pp. 51- 70. 16 Mirsky, A.F. and Grady, C.L., Towards the development of alternative treatments in absence epilepsy. In: M.S. Myslobodsky and A.F. Mirsky (Eds.), Elements of Petit Mal Epilepsy, Lang, New York, 1988, pp. 285-310. 17 Penry, J.K., Porter, R.J., and Dreifuss, F.E., Simultaneous recording of absence seizures with video tape and electroencephalography. A study of 374 seizures in 48 patients, Brain, 98 (1975) 427 - 440. 18 Porter, R.J., Penry, J.K. and Dreifuss, F.E., Responsiveness at the onset of spike-wave bursts, Electroenceph. Clin. NeurophysioL, 34 (1973) 239- 245. 19 Prechtl, H.F.R., Boeke, P.E., and Schut, T., The electroencephalogram and performance in epileptic patients, Neurology, 11 (1973) 296 - 302. 20 Richelle, M. and Lejeune, H., Timing competence and timing performance: a cross speciesapproach. In: J. Gibbon and L. Allan (Eds.), Timing and Time Perception, Annals of the New York Academy of Sciences, New York, 1984, pp. 254 - 268. 21 Schwab, R.S., Method of measuring consciousness in anacks of petit real epilepsy, Arch. Neurol. Psychiatry, 41 (1939) 215-217.
1 Aarts, J.H.P., Binnie C.D., Smit, A.M. and Wilkins, A.J., Selective cognitive impairment during focal and generalized epileptiform EEG activity, Brain, 107 (1984) 293 - 308. 2 Bentall, R.P., Lowe, C.F. and Beasty, A., The role of verbal behavior in human learning: II. Developmental differences, J. Exp. Anal. Behav., 43 (1985) 165 - 181. 3 Binnie, C.D., Kasteleijn-Nolst Trenit6, D.G.A., Smit, A.M. and Wilkins, A.J., Interactions of epileptiform EEG discharges and cognition, Epilepsy Res., 1 (1987) 239 - 245. 4 Bornstein, R.A., Pakalnis, A., Drake, M.E. and Suga, L.J., Effects of seizure type and waveform abnormality on memory and attention, Arch. Neurol., 45 (1988) 884- 887. 5 Brunia, C.H.M., Monosynaptic reflexes during generalized spike and wave activity, provoked by intermittent photic stimulation, Electroenceph. Clin. Neurophysiol., 49 (1980) 352-359. 6 Geller, M. and Geller A., Brief amnestic effects of spikewave discharges, Neurology, 20 (1970) 380-381. 7 Gibbs, F.A., Lennox, W.G. and Gibbs, E.L., The electroencephalogram in diagnosis and in localization of epileptic seizures, Arch. NeuroL Psychiatry, 36 (1936) 1225 - 1235. 8 Goldie, L. and Green, J.M., Spike and wave discharges and alterations of conscious awareness, Nature, 191 (1961) 200 - 201. 9 Hutt, S.J. and Fairweather, H., Information processing during two types of EEG activity, Electroenceph. Clin Neurophysiol., 39 (1973) 43- 51. 10 Hutt, S.J., Newton, J. and Fairweather, H., Choice reaction time and EEG activity in children with epilepsy, Neuropsychology, 15 (1977) 257- 267. II Hutt, S.J. and Gilbert, S., Effects of evoked spike-wave discharges upon short-term memory in patients with epilepsy, Cortex, 15 (1980) 445 - 457.
153 22 Sellden, U., Psychotechnical performance related to paroxysmal discharges in the EEG, Clin. Electroenceph., 2 (1971) 18-27. 23 Shimazono, Y., Hirai, T., Okuma, T., Fukuda, T. and Yamamasu, E., Disturbances of consciousness in petit mal epilepsy, Epilepsia, 2 (1953) 49 - 55. 24 Tizard, B. and Margerison, J.H., Psychological functions during wave-spike discharge, Br. J. Soc. Clin. Psychol., 3 (1963) 6 - 15. 25 van Luijtelaar, E.L.J.M. and Coenen A.M.L., Two types of electrocortical paroxysms in an inbred strain of rats,
Neurosci. Lett., 70 (1986) 393 - 397. 26 van Luijtelaar, E.L.J.M. and Coenen A.M.L., The WAG/Rij model for generalized absence seizures. In: J. Manelis, E. Bental, J.N. Loeber and F.E. Dreifuss (Eds.), Advances in Epileptology, Vol. 17, Raven Press, New York, 1989, pp. 7 8 - 8 3 . 27 van Luijtelaar, E.L.J.M., van de Werf, S.F., Vossen, J.M.H. and Coenen, A.M.L., Cognitive deficits and absence epilepsy in rats, Electroenceph. Clin. NeurophysioL, in press.
148
EPIRES 00418
Disturbances in time estimation during absence seizures in children
E.L.J.M. van Luijtelaar a, S.F.T.M. de Bruijn a, A.C. Declerck b, W.O. Renier c, J.M.H. Vossen a and A.M.L. Coenen a a Department of Psychology, University of Nijmegen, Nijmegen (The Netherlands), ~ Epilepsy Center Kempenhaeghe, Heeze (The Netherlands) and c Department of Child Neurology, University Hospital Nijmegen, Nijmegen (The Netherlands) (Received 28 January 1991; revision received 29 April 1991; accepted 13 May 1991)
Key words: Spike wave discharges; Absence epilepsy; Cognition; Performance; Timing behavior
In the present study children suffering from primary generalized absence epilepsy were asked to estimate time under EEG monitoring. They were asked to press a button when they thought that a fixed period of time had elapsed. Only the first response after the passage of the interval was reinforced. The dependent variable was the duration of the interval between the start of the trial and the first response, the post-reinforcement time. This parameter was used as an index for the accuracy of time estimation. The performance of the subjects in trials with and without spike wave discharges was compared. Short spike wave discharges ( < 3 s) prolonged the duration of the postreinforcement pause while longer ones reduced its duration. The prolongation was longer than could be anticipated from the duration of the spike wave discharge. It was also found that the time between the end of an EEG paroxysm and the first response was significantly shorter in trials with long spike wave discharges than in trials with short spike wave discharges. There were no differences between subjects with and without spike wave discharges. It was concluded that this type of time estimation task is sensitive in detecting cognitive disturbances induced by both short and long spike wave discharges. Moreover, it seems that after long spike wave discharges patients behave differently and are perhaps more severely disturbed. The duration of the misperception of the interval after the short spike wave discharges is larger than could be predicted from the duration of the paroxysms, which seems in contrast with the view that cognitive disturbances are only limited to the period of epileptic EEG phenomena.
INTRODUCTION
w i t h o u t evident changes in clinical b e h a v i o r 1. Brief lapses in the p a t i e n t ' s ability to m a i n t a i n c o n -
In m a n , the a b n o r m a l b r a i n activity characteristic o f
tact with the e n v i r o n m e n t are the p r i m a r y s y m p t o m o f
v a r i o u s types o f epilepsy is generally a c c o m p a n i e d by
generalized absence epilepsy. Various indications h a v e
an i m p a i r m e n t o f cognitive functioning. This i m p a i r m e n t a l processes to a c o m p l e t e arrest o f c o g n i t i v e ac-
been assigned to this s y m p t o m such as a loss o f c o n sciousness or a r e d u c t i o n in the ability to process i n f o r m a t i o n . Increases in reaction times 9A°,22, in omis-
tivity. T h e kind o f cognitive deficit is highly d e p e n d e n t
sions in c o n t i n u o u s p e r f o r m a n c e tasks as well as in
on the type, the d u r a t i o n a n d the n a t u r e o f the epileptic
stimulus detection are f r e q u e n t l y r e p o r t e d to a c c o m p a n y the characteristic spike w a v e discharges 15.
m e n t ranges f r o m a h a r d l y noticeable interference with
discharges and o f the r e q u i r e m e n t s o f the task 4. Subtle changes in m e n t a l f u n c t i o n i n g m a y even be present
It is agreed that s o m e tests are m o r e sensitive for c o g n i t i v e disturbances
Correspondence to: Dr. E.L.J.M. van Luijtelaar, Department of Psychology, University of Nijmegen, P.O. Box 9104, 6500 HE Nijmegen, The Netherlands.
t h a n others.
Simple m o t o r
tasks, such as r h y t h m i c t a p p i n g , simple r e a c t i o n t i m e or t r a c k i n g tasks are relatively little a f f e c t e d by generalized spike w a v e activity whereas m o r e c o m p l e x
0920-1211/91/$03.50 © 1991 Elsevier Science Publishers B.V. All rights reserved
149 tasks such as choice reaction time tasks, signal detection and short-term memory tasks are more sensitive 6,8,9,11,12,18,21,23,24. The sensitivity of tasks for cognitive impairments is especially important in the case of EEG discharges, since they often remain unnoticed by the patient or his environment. Furthermore, it is sometimes felt that cognitive disturbances are not present during short EEG paroxysms 12,19. In the present experiment, data will be presented which show that performance deficits do occur also during short-lasting spike wave activity. The task is a simple free operant learning task in which subjects had to estimate the duration of a fixed period of time. The response pattern induced by this fixed interval (FI) task is characterized by a long period without responding (the so-called post-reinforcement pause, initiated by the start of a new trial and ended by the first response of the subject in that trial). This task has been found to be sensitive for the presence of spike wave discharges in an animal model for absence epilepsy 27. METHODS
Subjects Fourteen subjects were recruited from 2 outpatient epilepsy clinics in the south-east of The Netherlands. All patients had a confirmed diagnosis of primary generalized absence epilepsy and their EEG had recently ( < 2 years) shown spontaneously occurring bilateral synchronous generalized 3-Hz spike wave activity with essentially normal interictal electrical activity. During the task hour, only 5 patients showed bilaterally appearing 3-Hz spike wave discharges. The ages of these 5 children ranged between 6 and 17 years, with a mean of 12. All patients were on mono- or polytherapy anti-epileptic drugs. Material The test device consisted of a multicolored wooden box (41 x 41 x 25 cm) containing a rotating disk (31.5 cm diameter), a non-transparent lid on top, and a push button. An opening in the front panel allowed collecting a token (pick-up tray). The box was made at the Instrumentation Department of the Psychology Department. The steerable slowly rotating disk in the box contained 8 compartments. A Playmobil toy could be placed in one of these compartments. Each compartment could fold down, and the toy would then fall
into the pick-up tray. An Apple II computer controlled the start and stop of the disk and controlled the presentation of the toy. During the experiment EEG traces (F0-C0, F4-C4, F3-C3, P3-O1, International 1 0 - 2 0 system) were registered, in which signals of 1 - 70 Hz were allowed to pass. The EEG was recorded on a cassette of a Medilog system, together with the responses of the subjects, the presentation of the reinforcers and the beginning and the end of the intervals. Later, all registrations were written out on paper.
Procedure After the EEG set-up was made and checked, subjects sat in front of the test box. They were shown the silently rotating disk, with its speed of 1 rpm, although this speed was not mentioned to the subjects. The toy was placed on the front compartment of the rotating disk. Subjects were shown that pressing the button had no consequences before the disk had made a complete turn, i.e., during the first minute. However, the first button press after completion of the turn stopped the rotating and the part of the disk on which the toy was placed folded downwards. The toy fell into the pick-up tray and could be collected by the subject. Next, the experimenter replaced the toy and a second trial was initiated by the experimenter who announced the start of the next trial by presenting an auditory stimulus. The dependent variable was the post-reinforcement pause, i.e., the interval between the auditory stimulus and the subject's first response. Subjects were allowed to get acquainted with the system, they were trained during 5 - l0 min. Next, the box was closed and the sign for the next trial was presented. Subjects were instructed to estimate the time of a complete turn of the disk. If they felt that this had happened, they pressed the button. In each trial the first response after the passage of 1 min was reinforced by the toy. Earlier bar presses had no consequences. The total session lasted 1 h. The collected toys were exchanged for Dfl 5, - in total at the end of the session for all subjects. The EEG was analyzed by an expert neurophysiologist who was unaware of the task requirements and behavioral data (A.D.). Each subject served as his/her own control, which means that possible differences between subjects with respect to medication are not critical, t-Tests for correlated means were used in order to test differences between
150 trials with and without spike wave discharges, the distribution free sign test was used to test differences in response latencies and a t-test for independent groups was used to test differences between subjects with and without spike wave discharges.
TABLE I The post-reinforcement pause (s) o f trials without, with short (< 3 s) and with long (> 3 s)spike wave discharges o f 5 children showing spike wave discharges in the EEG n
=
number of trials
Subject Without
n
Short
n
Long
V.M. S.H. J.S. C.B. M.B.
58.7 46.4 56.9 25.6 69.9
34 35 42 29 30
82.9 52.3 59.8 28.8 91.7
3 9 5 8 3
47.1 41.1 43.2 31.8 61.1
Mean S.D.
51.5 16.7
n
RESULTS Nine patients had no spike wave discharges during the experiment. The mean and standard deviation o f the post-reinforcement pause was 51.9 _+ 9.8 s, the number of responses per trial was 2.5 ± 0.9 and the number of trials per hour was 51.8 + 2.2. Five of the 14 subjects showed the generalized 3-Hz spike wave discharges in the E E G during the task. Their mean number of E E G paroxysms was 8.4 ± 3.1. The mean o f the number o f responses per trial was 2.6 ± 1.7 and the number o f trials per hour was 44.4 ± 5.3. There was no statistical difference in the number o f responses per trial between subjects with and without spike wave discharges. All other comparisons were within the subjects with spike wave discharges. A distinction between short ( < 3 s) and long ( > 3 s) spike wave discharges was made. Table I shows differential and opposite effects for trials with and without spike wave discharges. The post-reinforcement pause tended to be longer in trials with short spike wave discharges than in trials without spike wave activity (t -- 2.24, d f = 4, P < 0.10). A d ditionally, the post-reinforcement pause tended to be shorter after long spike wave discharges than after trials without spike wave activity (t = 2.46, d f = 4, P < 0.10). The difference between the post-reinforcement pause in trials with short and long spike wave discharges was significant (t -- 2.87, d f = 4, P < 0.05). F o r each subject the duration o f the E E G paroxysms was determined and c o m p a r e d with the prolongation o f the post-reinforcement pause. The prolongation o f the post-reinforcement pause (11.5 ± 10.6 s) was significantly longer (t = 4.24, d f = 4, P < 0.05) than the duration o f the spike wave activity (1.5 ± 0.6 s). Response latencies (time between the end o f the spike wave discharge and the first response) are presented in Table II. The non-parametric sign test showed that the latencies were significantly longer after short- than after long-lasting paroxysms (P < 0.05).
63.1 25.1
44.9* 10.7
* P < 0.05 (short vs. long).
TABLE II Response latencies (s) after short (< 3 s) and long (> 3 s) spike wave activity f o r the 5 children with spike wave discharges in the EEG during the test interval
t = time between the end of epileptic activity and the first response; n = number of spike wave discharges. Subject
t
t
short
n
long
n
V.M. S.H. J.S. C.B. M.B.
44.0 18.0 31.9 17.9 53.9
3 9 5 8 3
9.6 4.5 1.0 8.1 31.7
5 2 1 4 2
Mean S.D.
33.1 15.9
11.0" 12.1
* P < 0.05.
DISCUSSION All subjects p e r f o r m e d reasonably well in the fixed interval task. All, except the youngest patient, showed low-rate responding with only 1 - 2 responses in total at the end o f the interval. This behavioral pattern is c o m m o n l y found in h u m a n learning 2,14,2°. This means that, despite the lack o f extensive fixed interval training, the instruction and a short training were successful in inducing characteristic fixed interval responding. O f interest is also the similar number o f responses per trial o f the 5 subjects in trials without spike wave activity and o f the 9 subjects without any
151 sign of aberrant EEG activity. This suggests that interictal performance is not affected. Only 5 of the 14 subjects showed spike wave discharges during the experimental hour. It is likely that the mental activity during the task influenced and decreased the number of spike wave discharges 1. The duration of the post-reinforcement pauses shows that spike wave activity does affect the estimation of a time interval in all patients. Short spike wave discharges prolong the duration of the post-reinforcement pause and long spike-wave discharges reduce its duration. It is admitted that the distinction between short and long spike wave discharges is not solidly based on theoretical grounds. It would be preferable to make a distribution of the lengths of the spike wave discharges: a bimodal distribution with clear cut-off points would justify the distinction. However, in other studies the same distinction between short and long spike wave discharges has also been made 12. In some of them, cognitive aberrations were only present if the EEG paroxysm exceeded 3 s7'12. Only few papers can be found in which procedures or tasks are described which are sensitive for these short attacks 7,24. In this study, short spike wave discharges induce an increase of the post-reinforcement pause. Furthermore, this increase is larger than the duration of the spike wave discharge. It is often felt that the cognitive disturbances are limited to the period of the epileptic EEG phenomena and do not extend beyond the absence periods. Here, the amount of time that is wiped out or missed by the subjects is longer than the actual duration of the EEG paroxysm. It is not clear whether this 'wipe-out period' should be considered to be genuine amnesia, and if so, whether it is retro- or anterograde amnesia. Retrograde amnesia after discharges larger than 3 s has been described 6,12, whereas Hutt and Gilbert I 1 found retrograde amnesia in a short-term memory task. So far, anterograde memory deficits have not been reported. The present data do not allow us to draw conclusions with respect to putative memory deficits which extend the duration of the EEG signs. It is of interest that Mirsky and Grady 16 found a significant increase in the power of several EEG frequency bands as early as 20 s preceding the epileptic bursts, monotonically increasing till cumulating in the fully developed spike wave burst. This allows the possibility that the long cognitive deficits found in the present work are related to elec-
trophysiological antecedents preceding the fully developed spike wave discharge. In addition to this, Brunia 5 found indications for an altered motor system immediately after spike wave discharges. This latter result may indicate that antecedents of spike wave complexes are present after the EEG phenomena have disappeared. Long spike wave discharges induce an opposite effect on the post-reinforcement pause: it is shorter and relatively fast responding occurs after the paroxysmal activity has stopped. The response latency was significantly shorter than after short spike wave discharges. Thus it seems that after these long spike wave discharges a disorientation in time has taken place: the subjects seem to have no idea how much time has elapsed since the beginning of the trial and he/she responds quickly. It is well-known that after a long attack a patient's train of thought is often interrupted in such a way that he has to regain his thoughts after the attack. However, the patient will regularly carry on with his pre-ictal activities as if nothing has happened or may resume speaking or counting where he left off 17. This will occur with relatively simple tasks. Obviously, the time estimation task is sensitive to epileptic activity in such a way that a long attack interrupts the cognitive functioning so severely that disorientation and fast responding occurs after the attack. Although all subjects received the same instruction and sat in front of the same apparatus, it is not clear how subjects mastered this prospective time estimation task. Several strategies can be followed to accomplish this task: some subjects may count while others are really estimating the time that has elapsed and others may visualize the rotating of the disk. Although the type and nature of the cognitive strategy are of interest, the outcome of the present study is that similar effects with all 5 subjects were obtained, suggesting that the type of cognitive activity or the strategy used by the subjects did not have a major effect on performance disturbances induced by the spike wave activity. It is of interest that in rats with spontaneously occurring spike wave discharges the post-reinforcement pause in a fixed interval task was also prolonged in trials with spike wave discharges compared to trials without spike wave discharges 27. Also in agreement with the present study it was found that the prolongation of the post-reinforcement pause was larger than the duration of the EEG paroxysms. In the rat study,
152 a distinction between short and long spike wave
pected f r o m the duration o f the epileptic activity. Long
discharges was not made, but the agreement between
spike wave discharges disrupt ongoing activity even
the short-lasting spike wave discharges in m a n and rat
more severely, the subjects seem time disoriented.
is striking. This agreement strongly supports the validity o f this rat strain as a model for absence epilepsy in m an 25,26. It can be concluded that the timing task used in the
ACKNOWLEDGEMENTS We are indebted to Dr. J. Arends (Janssen Phar-
present experiment is sensitive in detecting cognitive
maceutics),
dysfunctioning.
In all subjects short spike wave
(Kempenhaeghe) for their cooperation in this study.
discharges prolonged the post-reinforcement pause
We also received assistance from the staff members of
and this prolongation exceeded the duration of the
the E E G Department of Kempenhaeghe, Mrs. Inge
E E G paroxysms. This relatively long 'missed' period
van der Linden and Mrs. Loan Kho, and secretarial
Dr.
L.
Oei
and
Dr.
J.
Doelman
is in contrast with the general opinion that cognitive
help f r o m Mrs. Tilly de Leijer. The study was made
disturbances are limited to the period of, in any case short, epileptic activity. We found that the magnitude
possible by support f r o m the Dutch Epilepsy Foundation 'D e Macht van bet Kleine'. Dr. E . L . J . M . van
o f the change in cognitive behavior after short spike
Luijtelaar is supported by the Royal Dutch A cad em y
wave paroxysms was much larger than could be ex-
of Sciences.
REFERENCES
12 Jus, A. and Jus, K., Retrograde amnesia in petit mal, Arch. Gen. Psychiatry, 6 (1962) 163 - 167. 13 Kasteleijn-Nolst Trenit6, D.G.A., Bakker, D.J., Binnie, C.D., Buerman, A. and van Raaij, M., Psychological effects of subclinical epileptiform EEG discharges: scholastic skills, Epilepsy Res., 2 (1988) l l l - 116. 14 Lowe, C.F., Determinants of human operant behavior. In: M.D. Zeiler and P. Harzem (Eds.), Advances in Analysis of Behavior." Vol. 1. Reinforcement and the Organisation of Behaviour, Wiley, Chichester, 1979, pp. 159- 192. 15 Mirsky, A.F., Information processing in petit mal epilepsy. In: B.P. Hermann and M. Seidenberg (Eds.), Childhood Epilepsies." Neurophysiological, Psychosocial and Intervention Aspects, Wiley, Chichester, 1989, pp. 51- 70. 16 Mirsky, A.F. and Grady, C.L., Towards the development of alternative treatments in absence epilepsy. In: M.S. Myslobodsky and A.F. Mirsky (Eds.), Elements of Petit Mal Epilepsy, Lang, New York, 1988, pp. 285-310. 17 Penry, J.K., Porter, R.J., and Dreifuss, F.E., Simultaneous recording of absence seizures with video tape and electroencephalography. A study of 374 seizures in 48 patients, Brain, 98 (1975) 427 - 440. 18 Porter, R.J., Penry, J.K. and Dreifuss, F.E., Responsiveness at the onset of spike-wave bursts, Electroenceph. Clin. NeurophysioL, 34 (1973) 239- 245. 19 Prechtl, H.F.R., Boeke, P.E., and Schut, T., The electroencephalogram and performance in epileptic patients, Neurology, 11 (1973) 296 - 302. 20 Richelle, M. and Lejeune, H., Timing competence and timing performance: a cross speciesapproach. In: J. Gibbon and L. Allan (Eds.), Timing and Time Perception, Annals of the New York Academy of Sciences, New York, 1984, pp. 254 - 268. 21 Schwab, R.S., Method of measuring consciousness in anacks of petit real epilepsy, Arch. Neurol. Psychiatry, 41 (1939) 215-217.
1 Aarts, J.H.P., Binnie C.D., Smit, A.M. and Wilkins, A.J., Selective cognitive impairment during focal and generalized epileptiform EEG activity, Brain, 107 (1984) 293 - 308. 2 Bentall, R.P., Lowe, C.F. and Beasty, A., The role of verbal behavior in human learning: II. Developmental differences, J. Exp. Anal. Behav., 43 (1985) 165 - 181. 3 Binnie, C.D., Kasteleijn-Nolst Trenit6, D.G.A., Smit, A.M. and Wilkins, A.J., Interactions of epileptiform EEG discharges and cognition, Epilepsy Res., 1 (1987) 239 - 245. 4 Bornstein, R.A., Pakalnis, A., Drake, M.E. and Suga, L.J., Effects of seizure type and waveform abnormality on memory and attention, Arch. Neurol., 45 (1988) 884- 887. 5 Brunia, C.H.M., Monosynaptic reflexes during generalized spike and wave activity, provoked by intermittent photic stimulation, Electroenceph. Clin. Neurophysiol., 49 (1980) 352-359. 6 Geller, M. and Geller A., Brief amnestic effects of spikewave discharges, Neurology, 20 (1970) 380-381. 7 Gibbs, F.A., Lennox, W.G. and Gibbs, E.L., The electroencephalogram in diagnosis and in localization of epileptic seizures, Arch. NeuroL Psychiatry, 36 (1936) 1225 - 1235. 8 Goldie, L. and Green, J.M., Spike and wave discharges and alterations of conscious awareness, Nature, 191 (1961) 200 - 201. 9 Hutt, S.J. and Fairweather, H., Information processing during two types of EEG activity, Electroenceph. Clin Neurophysiol., 39 (1973) 43- 51. 10 Hutt, S.J., Newton, J. and Fairweather, H., Choice reaction time and EEG activity in children with epilepsy, Neuropsychology, 15 (1977) 257- 267. II Hutt, S.J. and Gilbert, S., Effects of evoked spike-wave discharges upon short-term memory in patients with epilepsy, Cortex, 15 (1980) 445 - 457.
153 22 Sellden, U., Psychotechnical performance related to paroxysmal discharges in the EEG, Clin. Electroenceph., 2 (1971) 18-27. 23 Shimazono, Y., Hirai, T., Okuma, T., Fukuda, T. and Yamamasu, E., Disturbances of consciousness in petit mal epilepsy, Epilepsia, 2 (1953) 49 - 55. 24 Tizard, B. and Margerison, J.H., Psychological functions during wave-spike discharge, Br. J. Soc. Clin. Psychol., 3 (1963) 6 - 15. 25 van Luijtelaar, E.L.J.M. and Coenen A.M.L., Two types of electrocortical paroxysms in an inbred strain of rats,
Neurosci. Lett., 70 (1986) 393 - 397. 26 van Luijtelaar, E.L.J.M. and Coenen A.M.L., The WAG/Rij model for generalized absence seizures. In: J. Manelis, E. Bental, J.N. Loeber and F.E. Dreifuss (Eds.), Advances in Epileptology, Vol. 17, Raven Press, New York, 1989, pp. 7 8 - 8 3 . 27 van Luijtelaar, E.L.J.M., van de Werf, S.F., Vossen, J.M.H. and Coenen, A.M.L., Cognitive deficits and absence epilepsy in rats, Electroenceph. Clin. NeurophysioL, in press.